Re: [PATCH] [RFC] seqcount: Add lockdep functionality to seqcount/seqlock structures

[John Stultz <john.stultz@linaro.org>]

On 09/10/2013 01:55 AM, Peter Zijlstra wrote:
> On Mon, Sep 09, 2013 at 09:42:46PM -0700, John Stultz wrote:
>> @@ -38,10 +39,58 @@
>>   */
>>  typedef struct seqcount {
>>  	unsigned sequence;
>> +#ifdef CONFIG_DEBUG_LOCK_ALLOC
>> +	struct lockdep_map dep_map;
>> +#endif
>>  } seqcount_t;
>>  
>> -#define SEQCNT_ZERO { 0 }
>> -#define seqcount_init(x)	do { *(x) = (seqcount_t) SEQCNT_ZERO; } while (0)
>> +
>> +
>> +
>> +static inline void __seqcount_init(seqcount_t *s, const char *name,
>> +					  struct lock_class_key *key)
>> +{
>> +#ifdef CONFIG_DEBUG_LOCK_ALLOC
>> +	/*
>> +	 * Make sure we are not reinitializing a held lock:
>> +	 */
>> +	lockdep_init_map(&s->dep_map, name, key, 0);
>> +#endif
>> +	s->sequence = 0;
>> +}
>> +
>> +
>> +#ifdef CONFIG_DEBUG_LOCK_ALLOC
>> +# define SEQCOUNT_DEP_MAP_INIT(lockname) \
>> +		.dep_map = { .name = #lockname } \
>> +
>> +# define seqcount_init(s)				\
>> +	do {						\
>> +		static struct lock_class_key __key;	\
>> +		__seqcount_init((s), #s, &__key);	\
>> +	} while (0)
>> +
>> +static inline void seqcount_reader_access(const seqcount_t *s)
>> +{
>> +	seqcount_t *l = (seqcount_t *)s;
>> +	unsigned long flags;
>> +
>> +	preempt_disable();
>> +	local_irq_save(flags);
>> +	seqcount_acquire_read(&l->dep_map, 0, 0, _RET_IP_);
>> +	seqcount_release(&l->dep_map, 1, _RET_IP_);
>> +	local_irq_restore(flags);
>> +	preempt_enable();
>> +}
> Why the preempt and local_irq thing? Also preempt_disable is quite
> superfluous if you do local_irq_disable().

So since reader->writer recurision is ok, as is reader->reader
recurision, I wanted to avoid lockdep false positives if an irq lands in
between the aquire/release calls. So by doing the acquire/release w/
irqs off on the read side, we only trap the writer->reader recursion issues.

And agreed, the preempt_disable is overdoing it :)

Thanks for the review!
-john